Background: Surgical construction of an arteriovenous fistula is preferred for end-stage renal failure patients requiring long-term haemodialysis. Methods: Patients were randomised into two groups: brachial plexus group (n = 30) or local infiltration group (n = 30). In all patients, a radiocephalic arteriovenous fistula was created by an experienced surgeon using a standard surgical technique. In both groups 20 ml of 0.375% ropivacaine was used. Doppler assessment of vessels was performed at fixed time intervals. Results: Primary patency rate was 100% in the brachial plexus block group whereas there was 10% fistula failure rate in the local infiltration group (p-value = 0.237). Diameter of the vessels, peak systolic velocity, mean diastolic velocity, and blood flow at 30 minutes, 48 hours, 2 weeks, and 6 weeks after the fistula creation was significantly greater than the preoperative diameter in all patients (p-value < 0.05). Intergroup comparison revealed that vascular parameters were significantly better in the brachial plexus analgesia group versus local infiltration group at all observation points up to and including six weeks post fistula creation (p-value < 0.05). Conclusion: Brachial plexus anaesthesia significantly dilates the vessel diameter and increases blood flow whereas local infiltration has a negligible effect on vessel diameter and blood flow.